“Mastering FusionCatcher: A Complete Guide to RNA-Seq Fusion Detection” refers to the core concepts and operational workflows needed to utilize FusionCatcher, a premier, highly automated bioinformatics tool designed to identify somatic fusion genes, translocations, and chimeras from paired-end RNA-sequencing data. Primarily used in cancer research, FusionCatcher is engineered to deliver exceptional detection rates while heavily reducing false positives through multi-step alignment and extensive biological filtering. Key Capabilities of FusionCatcher
High Accuracy: Achieves competitive real-time PCR validation rates by distinguishing true somatic fusions from healthy germline events or artifacts.
Extreme Automation: Automatically handles adapter detection, quality trimming, and exon-exon junction building based on input read length.
Rigorous Filtering: Integrates massive curated databases (like COSMIC, TICdb, and ChimerDB) to flag and discard known false positives.
Functional Prediction: Analyzes the resulting transcript to predict its effect on the final protein structure (e.g., in-frame vs. out-of-frame). How the FusionCatcher Workflow Works
Behind the scenes, FusionCatcher processes raw sequencing reads through a sequence of data-heavy analysis tiers:
Preprocessing & Quality Control: Raw FASTQ files are evaluated, adapters are automatically stripped, and lower-quality sequences are trimmed.
Combination Alignment: To ensure no fusion junction is missed, the software orchestrates an alignment pipeline using Bowtie, STAR, and optionally BLAT.
Database Comparison: Candidate fusions are cross-referenced with genomic databases to filter out background noise, multi-copy genes, or read-throughs mapping to pseudogenes.
Evidence Consolidation: The tool scores the remaining pairs based on the number of unique reads spanning the exact fusion breakpoint. Technical Requirements & Setup
FusionCatcher is highly specialized and relies on strict structural dependencies: fusioncatcher/doc/manual.md at master – GitHub
Leave a Reply